Resistance to antineoplastic drugs is a major obstacle to the effectiveness of systemic therapy for cancer. Whereas the mechanisms of drug resistance are multiple and multifactorial, recent information suggests that the multidrug resistance (MDR) phenotype may be an important determinant of both experimental and clinical resistance to doxorubicin and other anthracyclines, to vinca alkaloids, and to etoposide. Studies in tissue culture and experimental animals have suggested the potential for reversal of MDR by a variety of drugs. We propose to relate the clinical outcome of anthracycline therapy to the presence of P-glycoprotein in cells of human tumors, and to investigate the possible reversal of the MDR phenotype using cyclosporin A. The specific aims of this proposal are: 1) to relate response to anthracycline therapy, to initial tumor levels of P-glycoprotein in patients with metastatic breast cancer; 2) to compare P-glycoprotein levels in tumor and normal tissue before and after anthracycline treatment; 3) to determine the maximally tolerated doses of cyclosporin A and doxorubicin used in combination, and to study their pharmacokinetics; and 4) to determine whether cyclosporin A can reverse doxorubin resistance in patients with metastatic colorectal or ovarian cancer. A phase II trial of anthracycline therapy in patients with metastatic carcinoma of the breast who have failed CMF and other non-anthracyclines containing first line treatment will be performed to assess aims 1 and 2. Patients with a measurable lesion amenable to biopsy scheduled to receive anthracyclines, will receive treatment after tissue biopsies have been taken for P-glycoprotein (as determined by immunoperoxidase staining) from tumor, white blood cells and buccal mucosa. Tumor response will be assessed and P-glycoprotein will be re-evaluated on cessation of treatment. A phase I trial to define the MTD and drug disposition of cyclosporin A and doxorubicin will be performed to address aim 3. Dose escalation of cyclosporin A given by infusion will be undertaken to reach the MTD followed by dose escalation of doxorubicin. On completion of the phase I trial, two phase II trials of doxorubicin and cyclosporin A will be performed in colorectal and ovarian cancer. Tumor response will be correlated with P-glycoprotein levels before and after treatment and with pharmacokinetic parameters. The proposed studies will evaluate the importance of P-glycoprotein in clinical resistance to MDR and determine whether cyclosporin A can reverse the MDR phenotype in patients.